Optimizing application throughput

Abstract

An approach is provided for optimizing application throughput. Data is collected about traffic flow in a computer network from one or more switches and one or more switches in the computer network. A hot spot destination in the computer network is identified by performing analytics on the collected data about the traffic flow against a traffic pattern in the computer network. An application-optimized software-defined networking (SDN) path is created for new flows to the hot spot destination. In response to an establishment of a new flow, the new flow is forwarded to the application-optimized SDN path.

What is claimed is: 
     
       1. A computer-implemented method of optimizing application throughput, the method comprising the steps of:
 generating, by one or more processors, an initial network design of a computer network that includes (i) an original network path to a backup server computer that includes a first set of links which are active links and (ii) a second set of links to the backup server computer that are inactive and redundant to the first set of links, the initial network design being based on a spanning tree protocol provided by processing in layer 2 that blocks the second set of links or an Internet Protocol (IP) route selection protocol provided by processing in layer 3 that shadows the second set of links, and the layer 2 and the layer 3 being layers in an Open Systems Interconnection model; 
 collecting, by the one or more processors, data about traffic flow in the computer network from one or more switches and one or more routers in the computer network; 
 identifying, by the one or more processors, the backup server computer as a hot spot destination in the computer network by performing analytics on the collected data about the traffic flow, the analytics indicating that the traffic flow matches a historical traffic pattern that indicates that (i) data is backed up from applications to the backup server computer and (ii) the applications from which the data is backed up are executed in a first pod of hosts; 
 in response to the backup server computer being identified as the hot spot destination, creating, by the one or more processors, an application-optimized software-defined networking (SDN) path that specifies a first path to the backup server computer and a second path to the backup server computer, the first and second paths utilizing the first set of links and the second set of links for new flows to backup data to the backup server computer in a subsequent network design of the computer network; 
 detecting, by the one or more processors, that (i) a first new flow is backing up first data from a first application to the backup server computer and (ii) the first application is being executed by a first host in the first pod; 
 detecting, by the one or more processors, that (i) a second new flow is backing up second data from a second application to the backup server computer and (ii) the second application is being executed by a second host in the first pod; 
 in response to the detecting that the first and second new flows are backing up the first and second data, respectively, to the backup server computer which is the hot spot destination and that the first and second applications are being executed in the first pod, replacing, by the one or more processors, the initial network design with the subsequent network design and activating, by the one or more processors, the second set of links in the subsequent network design; 
 forwarding, by the one or more processors, the first and second new flows to the application-optimized SDN path so that the first new flow is forwarded to the backup server computer via the first set of links and the second new flow is forwarded to the backup server computer via the second set of links that are active but were inactive prior to the initial network design being replaced with the subsequent network design; and 
 in response to the replacing the initial network design with the subsequent network design and the forwarding the first and second new flows, bypassing the processing in layer 2 that blocks the second set of links, bypassing the processing in layer 3 that shadows the second set of links, and providing a throughput of the backing up of the first and second data by the first and second applications using the first and second set of links according to the subsequent network design, so that the throughput is improved over another throughput of a backup of the first and second data by the first and second applications using the first set of links but not the second set of links according to the initial network design. 
 
     
     
       2. The method of  claim 1 , further comprising the step of configuring, by the one or more processors, the application-optimized SDN path on SDN switches. 
     
     
       3. The method of  claim 1 , further comprising the step of:
 providing at least one support service for at least one of creating, integrating, hosting, maintaining, and deploying computer readable program code in the computer, the program code being executed by a processor of the computer to implement the steps of generating the initial network design, collecting the data about the traffic flow, identifying the backup server computer as the hot spot destination, creating the application-optimized SDN path, detecting that (i) the first new flow is backing up the first data and (ii) the first application is being executed by a first host in the first pod, detecting that (i) the second new flow is backup the second data and (ii) the second application is being executed by a second host in the first pod, replacing the initial network design with the subsequent network design, activating the second set of links, and forwarding the first and second new flows to the application-optimized SDN path, bypassing the processing in layer 2, bypassing the processing in layer 3, and providing the throughput so that the throughput is improved over another throughput. 
 
     
     
       4. A computer program product for optimizing application throughput, the computer program product comprising a computer readable storage medium having computer readable program code stored on the computer readable storage medium, wherein the computer readable storage medium is not a transitory signal per se, the computer readable program code being executed by a central processing unit (CPU) of a computer system to cause the computer system to perform a method comprising the steps of:
 the computer system generating an initial network design of a computer network that includes (i) an original network path to a backup server computer that includes a first set of links which are active links and (ii) a second set of links to the backup server computer that are inactive and redundant to the first set of links, the initial network design being based on a spanning tree protocol provided by processing in layer 2 that blocks the second set of links or an Internet Protocol (IP) route selection protocol provided by processing in layer 3 that shadows the second set of links, and the layer 2 and the layer 3 being layers in an Open Systems Interconnection model; 
 the computer system collecting data about traffic flow in the computer network from one or more switches and one or more routers in the computer network; 
 the computer system identifying the backup server computer as a hot spot destination in the computer network by performing analytics on the collected data about the traffic flow, the analytics indicating that the traffic flow matches a historical traffic pattern that indicates that (i) data is backed up from applications to the backup server computer and (ii) the applications from which the data is backed up are executed in a first pod of hosts; 
 in response to the backup server computer being identified as the hot spot destination, the computer system creating an application-optimized software-defined networking (SDN) path that specifies a first path to the backup server computer and a second path to the backup server computer, the first and second paths utilizing the first set of links and the second set of links for new flows to backup data to the backup server computer in a subsequent network design of the computer network; 
 the computer system detecting that (i) a first new flow is backing up first data from a first application to the backup server computer and (ii) the first application is being executed by a first host in the first pod; 
 the computer system detecting that (i) a second new flow is backing up second data from a second application to the backup server computer and (ii) the second application is being executed by a second host in the first pod; 
 in response to the detecting that the first and second new flows are backing up the first and second data, respectively, to the backup server computer which is the hot spot destination and that the first and second applications are being executed in the first pod, the computer system replacing the initial network design with the subsequent network design and the computer system activating the second set of links in the subsequent network design; 
 the computer system forwarding the first and second new flows to the application-optimized SDN path so that the first new flow is forwarded to the backup server computer via the first set of links and the second new flow is forwarded to the backup server computer via the second set of links that are active but were inactive prior to the initial network design being replaced with the subsequent network design; and 
 in response to the replacing the initial network design with the subsequent network design and the forwarding the first and second new flows, the computer system bypassing the processing in layer 2 that blocks the second set of links, the computer system bypassing the processing in layer 3 that shadows the second set of links, and the computer system providing a throughput of the backing up of the first and second data by the first and second applications using the first and second set of links according to the subsequent network design, so that the throughput is improved over another throughput of a backup of the first and second data by the first and second applications using the first set of links but not the second set of links according to the initial network design. 
 
     
     
       5. A computer system comprising:
 a central processing unit (CPU); 
 a memory coupled to the CPU; and 
 a computer readable storage medium coupled to the CPU, the computer readable storage medium containing instructions that are executed by the CPU via the memory to implement a method of optimizing application throughput, the method comprising the steps of:
 the computer system generating an initial network design of a computer network that includes (i) an original network path to a backup server computer that includes a first set of links which are active links and (ii) a second set of links to the backup server computer that are inactive and redundant to the first set of links, the initial network design being based on a spanning tree protocol provided by processing in layer 2 that blocks the second set of links or an Internet Protocol (IP) route selection protocol provided by processing in layer 3 that shadows the second set of links, and the layer 2 and the layer 3 being layers in an Open Systems Interconnection model; 
 the computer system collecting data about traffic flow in the computer network from one or more switches and one or more routers in the computer network; 
 the computer system identifying the backup server computer as a hot spot destination in the computer network by performing analytics on the collected data about the traffic flow, the analytics indicating that the traffic flow matches a historical traffic pattern that indicates that (i) data is backed up from applications to the backup server computer and (ii) the applications from which the data is backed up are executed in a first pod of hosts; 
 in response to the backup server computer being identified as the hot spot destination, the computer system creating an application-optimized software-defined networking (SDN) path that specifies a first path to the backup server computer and a second path to the backup server computer, the first and second paths utilizing the first set of links and the second set of links for new flows to backup data to the backup server computer in a subsequent network design of the computer network; 
 the computer system detecting that (i) a first new flow is backing up first data from a first application to the backup server computer and (ii) the first application is being executed by a first host in the first pod; 
 the computer system detecting that (i) a second new flow is backing up second data from a second application to the backup server computer and (ii) the second application is being executed by a second host in the first pod; 
 in response to the detecting that the first and second new flows are backing up the first and second data, respectively, to the backup server computer which is the hot spot destination and that the first and second applications are being executed in the first pod, the computer system replacing the initial network design with the subsequent network design and the computer system activating the second set of links in the subsequent network design; 
 the computer system forwarding the first and second new flows to the application-optimized SDN path so that the first new flow is forwarded to the backup server computer via the first set of links and the second new flow is forwarded to the backup server computer via the second set of links that are active but were inactive prior to the initial network design being replaced with the subsequent network design; and 
 in response to the replacing the initial network design with the subsequent network design and the forwarding the first and second new flows, the computer system bypassing the processing in layer 2 that blocks the second set of links, the computer system bypassing the processing in layer 3 that shadows the second set of links, and the computer system providing a throughput of the backing up of the first and second data by the first and second applications using the first and second set of links according to the subsequent network design, so that the throughput is improved over another throughput of a backup of the first and second data by the first and second applications using the first set of links but not the second set of links according to the initial network design.